Hybrid CNN and RNN for Twitter Sentiment Analysis

Online social media networks have developed into a widespread and significant platform for people to freely express their thoughts and emotions in this era of digital communication. This change in communication habits has important ramifications for brands looking to connect with and understand their target market. Social media is one of the biggest sources of unstructured data, but it takes time to analyze it and extract its meaning. This paper attempts to forecast sentimental analysis based on Twitter user’s tweets. We used a Twitter dataset which contains around 520 k Tweets, to predict whether the tweet has a positive or a negative connotation which can further help us to understand the mental state of the user. Then, using this dataset, we develop different deep learning models (including RNN and CNN). We examine the outcomes of applying recurrent neural network (RNN) and convolutional neural network (CNN) to these models. Finally, we suggested an approach that combines RNN and CNN to fully exploit each technology’s advantages: RNN can learn temporal and context features, notably long-term dependency between multiple entities, while CNN is capable of catching numerous potential features. The result demonstrates that our method is superior to the majority of the existing methods.